Oxygen uptake by the pulmonary circulation is a chemical reaction. The physicochemical attributes of oxygen are critical when studying pulmonary oxygen toxicity. Extent of lung injury depends on the percentage of oxygen in an oxygen:nitrogen mix in polybaric circumstances . Further change in extent of lesion follows when other gases are used in the inhalant mix instead of nitrogen (Shanklin and Lester, 1972), with oxygen at 21-100% of the mix. Comparative subatmospheric oxygen levels down to 3% in hydrogen, helium, nitrogen, argon, or sulfur hexafluoride, were run with and without ventilatory distress by the Farber (1937) model, bilateral cervical vagotomy (BCV). This yielded coherent results indicating a need to consider molecular characteristics at the atomic level. Molecular mass and size, gas viscosity, and thermal conductivity yielded no obvious correlates to lung injury. Saturation of the outer electron shells of the diluents fit the empiric data, prospectively an interaction between oxygen and nitrogen from their electronegativity and closely approximate molecular mass, size, and shape. The lesion is essentially eliminated at 7% oxygen in nitrogen. At 3% oxygen, the least lesion is found with N 2 , H 2 , and SF 6 , all gases with incomplete outer electron shells, allowing for transient, possibly polarized, covalent bonding with oxygen as the significant minority component in the mix. Argon and helium do not interfere with oxygen. With 3% oxygen in argon without BCV, the experiments ran so long (>70 hours) they were terminated once the point had been made. 3% oxygen in argon after BCV yielded a mean survival more than twice that of BCV in air, indicating a remarkable degree of nitrogen interference with oxygen in the respiratory medium of terrestrial animal life. Argon displayed other advantages for the lung compared to nitrogen. Hydrogen, nitrogen, and oxygen are diatomic molecules, a feature which does relate to the extent of lung injury, but only oxygen is paramagnetic.Magnetic effects on lesion formation were tested: [1] with ventilatory distress induced in newborn rabbits, and [2] in young adult female white mice exposed to 100% oxygen without added mechanical distress. A noninvasive model for ventilatory distress, thoracic restraint (TR), with longer mean survivals of 40-50 hours, was employed rather than the Farber model. Parallel runs with TR, one subset receiving 100% oxygen in a plastic chamber resting on six strong ring magnets with measured fields up to +1200 gauss, the other plain 100% oxygen, were performed. Both subsets developed moderate metabolic acidosis with average weight losses circa 25%, but over different time courses, 82.89 ± 4.91 hours in magnetized oxygen, 55.4 per cent longer than the 53.34 ± 9.82 hours in plain oxygen ( p <0.001). The longer survival in magnetized oxygen meant extensive lung injury (99.57 ± 0.42% pleural surface, versus 83.86 ± 14.03%), but the rate of lesion formation was 30.89 per cent faster in plain oxygen (1.5722% per hour) than in magnetized oxygen...